Sound familiar? The ad platform shows that paid search won the deal. The sales rep is certain it was the conversation at a trade show six months earlier. Both are right, and both are wrong. That is exactly the problem with last-click attribution in B2B.
Last-click assigns 100% of the credit to the final touchpoint before a lead converts. In an e-commerce context with a purchase journey measured in minutes, that's a reasonable approximation. In a B2B sales cycle that often spans many months and, according to GartnerSourceGartner: The B2B Buying JourneyGartner for SalesGartner describes the typical buying group for a complex B2B solution as involving 6-10 decision-makers.View source ↗, typically involves 6-10 decision-makers with different roles, budgets, and risk profiles, it's a systematic measurement error - and that error costs you budget every single quarter.
01Why Last-Click Systematically Lies in B2B
The mechanism is simple, but the consequences aren't. When a buyer finally converts, it's almost always via a branded search or a direct visit - because they've already decided and just need to find their way back to your website. Last-click sees that search and concludes "branded search" drove the deal. In reality, it was the awareness campaign eight months earlier, the LinkedIn post a colleague shared, or the webinar another stakeholder attended that actually started the decision process.
Add to that the fact that B2B buying journeys largely happen untracked: conversations in Slack communities, LinkedIn DMs, word-of-mouth between colleagues in the same industry, and internal discussions in the buying committee you never see. Your analytics platform only measures the touchpoints it can see - and the touchpoints it can see are systematically skewed toward the bottom of the funnel.
02The Three Attribution Failures That Keep Recurring
- 1
First failure
Budget flows toward the bottom of the funnel because it "shows ROI" - while top-funnel awareness, which actually starts the pipeline, gets cut because it "doesn't convert" in a model that measures it wrong.
- 2
Second failure
Marketing and sales fight over credit for every closed deal because neither trusts the data.
That drains energy that should go toward improving the shared pipeline, and it undermines marketing's credibility with leadership - precisely when they need it most.
- 3
Third failure
Decision-makers optimise towards a number that doesn't reflect reality, and systematically make the wrong budget decisions - often for years, without anyone noticing, because the number looks precise. Precision is not the same as accuracy.
03A Measurement Model That Follows the Buying Group, Not the Click
The fix isn't "better tracking." It's a different model. Three components that together produce an accurate picture:
Data-driven multi-touch across the full funnel
Instead of assigning 100% of credit to one touchpoint, a data-driven model distributes credit algorithmically based on each touchpoint's actual correlation with conversion - not an arbitrary rule like "40% to first, 40% to last, 20% in the middle." It requires clean, consistent data across channels, but it's the only model that isn't built on an assumption you never tested. If you close few deals a year, though, the data is too thin for an algorithm, and a simple, jointly agreed allocation rule is a better place to start.
Account-level, not user-level
A single B2B conversion typically involves multiple people from the same company interacting with different channels at different times. Measure at the individual user-journey level and you lose sight of the fact that the CFO saw a LinkedIn ad while the IT lead downloaded a whitepaper and the procurement manager was the one who filled out the form. Account-level attribution stitches these journeys into one coherent story. In practice, that means every contact is linked to their company in the CRM, with source and campaigns stored on each contact. You can then see which channels touched the companies that became customers - not just the person who filled out the form.
Sales-sourced influence tagging
No tracking technology captures a conference meeting or a referral from a former colleague. The fix is low-tech: have your account executives systematically log which touchpoints and content buying committee members themselves mention in sales conversations. It's not perfect data, but it's data that closes the blind spot no pixel can see.
04How to Rebuild Your Attribution Model in Practice
- 1
Step 1
Audit your current model for bias.
Run a simple test - compare how much budget goes to top-, mid-, and bottom-funnel today against your average sales cycle length. If 80% of budget goes to bottom-funnel channels in a 12-month sales cycle, you already have your answer.
- 2
Step 2
Implement account-based multi-touch.
This requires connecting the CRM to the ad platforms (e.g., via offline conversion import in Google Ads) and to your analytics, and grouping contacts at company level in the CRM, so you can follow an account - not just a user - from first touchpoint to closed deal. GA4 measures at user and session level and can't attribute at account level on its own - that layer has to be built in the CRM or a dedicated tool. Without that connection, you're still only measuring a fraction of the journey.
- 3
Step 3
Layer a qualitative signal on top.
Build a simple process where sales logs "influencing touchpoints" in the CRM at every closed deal - 2 minutes per deal that, over time, reveals patterns no platform can give you alone.
A typical pattern once the new model is in place: a markedly larger share of real pipeline value can be traced back to top-funnel activities than the modest share last-click gave them. That is often enough for budget to be reallocated.
05Do You Need to Buy a New System?
Rarely as the first step. Most B2B companies already have the most important tool: the CRM. Build in three layers, so the investment follows the need:
Use the CRM you have
Link contacts to companies, store source and campaign on each contact, and ask "How did you hear about us?" both in forms and in sales' first meeting. It needs no new licences, only discipline.
Connect the CRM to the ad platforms
Send qualified leads and won deals back to platforms such as Google Ads and LinkedIn, so the platforms optimise for pipeline rather than form fills. It is typically an integration, not a new system.
A dedicated attribution tool when volume calls for it
Platforms such as Dreamdata or HockeyStack are built for account-level B2B attribution and bring together data from website, ads and CRM. They add the most value when there are many channels, several markets and enough deals for patterns to become clear. With few deals a year, layers 1 and 2 usually deliver most of the value.
The order is the point: the tool comes last. Without clean data and a shared definition of what counts, even the most expensive platform just gives you a more sophisticated wrong number.
06The Political Reality Behind the Attribution Model
The real barrier is rarely technical. Marketing owns the awareness budget and wants credit for pipeline. Sales owns the relationships and wants credit for closed deals. Finance wants one simple ROI number to report to the board. Changing the attribution model is therefore as much an organisational negotiation as a technical implementation - and that's exactly where outside advisory creates the most value: a party with no internal stake in who gets the credit, but a stake in the number being right.
This model only works if your CRM data is clean and your conversion tracking is already accurate - if you're still fighting basic tracking integrity, that's where to start before building an attribution layer on top.
The takeaway
Attribution isn't a reporting exercise. It's the decision basis for where the next several million in marketing budget goes. It deserves a model that actually reflects how your customers buy.